Democratic Leaders Want An Affordability Debate On Ai. Critics Say They’re Ducking The Real Fight.

House Democrats are heading into the midterms with a pocketbook-focused message on artificial intelligence — one narrowly centered on the energy costs of data centers, but far milder than the populist outcry against Big Tech that is stymieing AI projects across the country.

The theme that party leaders are settling on — articulated by multiple House Democrats who spoke to POLITICO — meshes with Democrats’ wider attacks on soaring fuel prices and other affordability flashpoints of the second Trump era. It also avoids a direct collision between Democrats and the AI industry, which has amassed hundreds of millions of dollars it could spend on this year’s elections.

The thrust of the message: The U.S. needs to win the race to dominate AI. But ordinary Americans shouldn’t bear the burden of supplying data centers’ voracious energy needs.

But critics say it risks placing Democrats out of step with growing public alarm about AI’s potential to wipe out jobs, eradicate privacy and possibly even endanger humanity. Those fears have prompted some progressives to call for a temporary ban on building data centers, along with an expanded social safety net to combat potential job loss.

Already, local governments and voters from Wisconsin to Missouri to the Phoenix suburbs have lashed out against big AI data center projects, posing a major obstacle to the industry’s efforts to meet rapidly growing demand.

“It’s very clear that many Democrats are missing the boat on a clear proof point that they can fight corporate power, and they’re missing it once again,” said Tommy McDonald, a strategist at the progressive consulting firm Fight Agency, which launched last year with the aim of electing anti-establishment candidates who can appeal to the working class.

The complaint is the latest sign of Democrats’ divide on how to respond to AI, even as the party hopes to reclaim control of both chambers of Congress.

House Democrats in potentially tough races in November argue that they’re offering the kind of direct, clear messaging that voters need to hear — on an issue many people are still making up their minds about.

“Leadership has been focused entirely on costs,” said Rep. Greg Landsman (D-Ohio), who will face a competitive challenge in a redrawn district in November, when asked whether he has received any guidance from Democratic leaders on how to talk about AI. “That’s the most urgent issue, is making sure that we are protecting households, American ratepayers, from the increased cost associated with these data centers.”

‘Summoning the demon’

His comments echo a message last week from House Minority Leader Hakeem Jeffries (D-N.Y.), who said in response to a question from POLITICO that “we have to protect the American homeowner. We have to protect the American ratepayer from some of the downsides of the explosion of what has been taking place across the country, particularly in connection with data centers.”

At the same time, Jeffries said Democrats want to ensure that “our homegrown companies can continue to lead the world in this transformative technology.”

That stance bears some similarities to President Donald Trump’s pledge this year to shield ordinary Americans from AI’s energy costs, in part by forcing many data centers to pay for their own power supplies.

But it’s miles away from proposals by some progressives to block data center construction altogether. Those include legislation championed by Sen. Bernie Sanders (I-Vt.) and Rep. Alexandria Ocasio-Cortez (D-N.Y.) that would prohibit building or expanding the AI hubs until Congress passes laws to ensure the technology is safe, won’t worsen consumers’ power bills or harm the environment and “will benefit workers, not just the wealthy owners of Big Tech companies.”

Sanders’ bill cites warnings about AI from various tech leaders, including Elon Musk’s prediction last year that “AI and robots will replace all jobs.” It also quotes Musk’s remark from more than a decade ago that “with artificial intelligence we are summoning the demon.”

The proposal has gotten a dim reception from other Democrats in Congress. Sen. Mark Warner (D-Va.) dismissed the idea of a moratorium as “idiocy” during a recent event hosted by Axios, saying it would mean that “China is going to move quicker.”

Warner joined with Sen. Ted Budd (R-N.C.) on Thursday in one bipartisan approach to AI, introducing a bill that would track how the technology is reshaping the workforce. The same week, House Democratic Caucus Vice Chair Ted Lieu (D-Calif.) and Rep. Jay Obernolte (R-Calif.) introduced legislation that would formalize recommendations of a bipartisan AI task force, including by improving standards for the technology, modernizing federal risk management, investing in workforce development programs and deterring harmful deepfakes.

Lieu’s message is in line with much of his party’s leadership: “What we want to do is make sure that we continue to innovate in AI, but also have reasonable guardrails to make sure it doesn’t harm Americans.”

But for Democrats who want the party to take a more aggressive approach to AI regulation, specifically as it relates to potential job loss, a focus solely on data centers’ energy costs obfuscates a winning message about corporate overreach and Silicon Valley overlords. For others, any talk of tougher AI regulation can invite unwanted industry spending in tight races.

A gusher of tech money

In interviews with over a dozen political strategists and consultants working on AI issues, two countervailing narratives emerged — that Democrats should go harder in attacking the Trump administration and many conservatives on their AI policy, but that doing so could bring a wave of pro-AI super PAC cash aimed at defeating them.

“There’s a large contingent within the Democratic Party that maintains a posture that Democrats won when they had the backing of Silicon Valley, and we need to get them back,” said a Democratic strategist heavily involved in AI issues, who was granted anonymity to speak candidly about strategy. “Not only is that objectively misguided, but … a strategy of appeasement is unlikely to be successful. If we want Silicon Valley to pay attention to us, they need to be scared of us and believe there’s a real chance regulation is coming.”

As a political issue, AI regulation may seem to be a winner: A 46 percent plurality of Americans say they would prioritize ensuring that AI is safe and well-regulated even if that means China develops it faster, according to April results from The POLITICO Poll. This was true of 47 percent of Republicans and 57 percent of Democrats.

Americans in the April survey also supported the government imposing strict regulations on AI companies (38 percent) or setting broad principles (29 percent). Just 13 percent said the government “should stay out of it and let the market decide.”

And yet, many Democrats without a background in tech appear content to avoid being drawn into a messy, big-spending political battle over AI, one that’s increasingly shaping up as a proxy fight between competing super PACs aligned with the leaders of the industry’s two most prominent companies.

The pro-industry super PAC Leading the Future has more than $100 million in the bank, much of it from executives at companies including OpenAI, the maker of ChatGPT. It is spending big against candidates who favor tough regulations on AI, most prominently in an effort to defeat Democratic New York state lawmaker Alex Bores’ campaign for a U.S. House seat.

A more pro-regulation wing of the debate — often described by the label “AI safety” — has the financial backing of Public First, a dueling super PAC that has received significant donations from executives at OpenAI’s archrival Anthropic. Anthropic has said it is concerned about the potential catastrophic risks to humanity of advanced AI systems, a posture that has sometimes brought it into coalition with more pro-regulation members of Congress.

“Leading the Future wants to move fast and lock in outcomes before voters have formed real opinions,” said former Rep. Brad Carson (D-Okla.), the founder of Public First. His group opposes any efforts to preempt state or local legislation on AI with what he sees as an overly loose federal standard.

“When we talk to Leader Jeffries and other Democrats, we’re not asking anyone to become an overnight AI expert or to make them walk point on AI policy,” Carson said. “We want them to follow the intuitions of their constituents, develop their own opinions on their own timeline, and not get boxed in before they’re ready. Our job is to protect that ability.”

When reached for comment, Leading the Future co-head Josh Vlasto said Public First was intimidating members by insisting they support Anthropic’s policy priorities.

“We will continue to support policymakers and candidates who will work to pass a national regulatory framework that creates jobs for American workers, wins the race against China and protects the safety of kids, users and communities,” Vlasto said.

‘If you do this, this is who will try to destroy you’

The dueling PACs, never missing a chance to snipe at each other, are still searching for champions of their agendas. Meanwhile, Democratic House leadership is advising vulnerable Democrats on the possibility that they could face an onslaught of cash spent against them should they choose to attack the AI industry, according to three people familiar with the guidance who were granted anonymity to discuss internal strategic conversations.

“From past experience with the policy teams at the [Democratic Congressional Campaign Committee], a lot of what they do is give road maps — ‘if you do this, this is who will try to destroy you.’ It’s sort of a choose your own adventure,” said a fourth person, a swing state Democrat who was granted anonymity to candidly discuss internal party politics. The person said AI “is evolving so fast that everyone is circling each other trying to figure out who’s a potential threat and who’s a potential foe.”

Spokespeople for the DCCC and Jeffries declined to comment.

Both parties are making a political mistake by not being bolder and louder with their concerns about AI safety, said Brendan Steinhauser, a conservative consultant and the CEO of the Alliance for Secure AI, a think tank focused on potential harms from AI development.

“I think that both parties need to do more, take a stronger stance and get out in front of this issue in terms of the jobs issue and the national security side,” Steinhauser said. He added: “What I’ve told my friends who are consultants — don’t do what I’m asking you to do, just look at the data. Look at the polling. Look at what the data says about how to talk about AI.”

Democrats have fallen on the wrong side of Big Tech spenders before: Then-Sen. Sherrod Brown (D-Ohio) lost his seat in 2024 after a super PAC funded by cryptocurrency companies spent more than $40 million to defeat him. Many of the same people who successfully defeated Brown are now involved with running Leading the Future.

But some within the party believe that AI is a different beast, because voters are quickly codifying strong opinions on the issue in a way they haven’t on crypto.

"Crypto was abstract," said Democratic strategist Jesse Ferguson, a former director at the DCCC who now advises several nonprofits and super PACs. "AI is personal — your job, your kids, your privacy … you can't outspend fear. And right now, voters fear unchecked AI more than they fear any super PAC.

"For Democrats, it's an opportunity big enough to drive a Mack truck through, as long as you’re willing to turn the ignition.”